Sunday, September 06, 2009

Going behind the horserace numbers

We had two polls out last week, one from Angus Reid and one from Ekos, both of which showed the Liberals and Conservatives neck-in-neck as we go into the fall session of parliament and a possible election campaign.

I don’t put much import in the polls, other than affirming that Ipsos poll looks to be the 19th time out of 20, that anything is still possible and that, as always, the campaign will be determinative. I would say it’s a positive for the Liberals to be tied coming out of a summer when the opposition always has a hard time attracting national attention, particularly when the government has billions of dollars in stimulus funding to announce, and re-announce. It will be interesting to see where the numbers go now that people are paying attention once again.

Rather than focus totally on national horse race numbers though, I wanted to dive a little deeper and look at some of the other numbers in the two polls that I found interesting.

Regional highlights

When I look at regional numbers, I’m usually most interested in Ontario, Quebec and BC. The first two for strategic reasons, the latter personal.

Ekos has the Liberals restoring a decent lead in Ontario, at LPC 40.4, CPC 33.1, NDP 15.6. The Angus Reid numbers have it a bit tighter, at LPC 40, CPC 37, NDP 14. The Liberals need to be at 40 per cent or higher in Ontario to be competitive nationally. The NDP number is also important, because higher NDP numbers will create favourable vote splits for the Conservatives. The NDP have shown some weakness in Ontario lately, but the Liberals would like to see more daylight between them and the Conservatives.

Ekos also provides some interesting metro numbers, although with slightly higher margins of errors. In Toronto, Ekos has it Liberals 44.8, CPC 35.4, NDP 11.8, GPC 8.1 with a 6.6 MOE. The CPC shows strong strength in the GTA here but what surprised me was the low number for the NDP. That could make things interesting for a few of their incumbents.

In Quebec. Angus Reid has it BQ 35, LPC 31, CPC 18, NDP 10. Ekos calls it BQ 32.3, LPC 30.9, CPC 19.4, NDP 9.8. Again, two sets of similar numbers. Continued CPC weakness in Quebec will cost them seats they’ll need to make up elsewhere just to tread water. The BQ and the Liberals will gain, probably the BQ a little more. The NDP numbers are just above the Green Party in Quebec; if I was Thomas Mulcair I’d be a tad concerned.

Ekos puts Montreal at LPC 32.7, BQ 32.3. CPC 18.2, NDP 9.1 with a 6.5 MOE.

Finally, out in BC, Ekos has it CPC 35.3, NDP 25.5, LPC 24.9, GPC 14.3. At Angus Reid, its CPC 34, NDP 33, LPC 24, GPC 9. There’s been a lot of fluidity in the BC numbers of late. I don’t see the tightness of the Angus Reid number, but I also don’t see the high Green numbers that Ekos has holding. The Greens polled strongly in BC during the last campaign too, but came back down by e-day.

The thing to remember about BC is that the Conservatives largely run the table in rural BC, and compete just with the NDP in coastal BC and a few interior and Island ridings. So large Conservative rural pluralities can create a mini-Alberta effect in the provincial numbers. Liberal strength is largely concentrated in the Lower Mainland and South Island. In Vancouver, Ekos puts it at LPC 32.4, CPC 30.7, NDP 26.5, GPC 10.4 (MOE 10.5). While there’s a few strong CPC ridings in that mix, those numbers are still a little tighter than I’d like.

Doing the demographics

But enough of horserace numbers. What really interests me is when the pollsters make demographic data available. For example, support by gender.

The gender breakdowns from Ekos offer some interesting insights. Amongst men, its CPC 35.8 to LPC 33.1. That’s tighter than I recall it being for awhile, usually the CPC leads strongly with men. Among women, its LPC 32.2 to CPC 29.5. The Liberals have traditionally done very well with female voters; they lost that edge last election to the Conservatives and it was a big factor in the poor results. We look to be getting women back, and that’s a very positive sign. But we need to continue to build our support there with targeted policies and messaging to be competitive for government in the next election.

This is borne out by the Ontario numbers. Here, they show the seven point Liberal lead in Ontario is largely thanks to female voters, who preferred the Liberals 41.1 to 27.9 for the Conservatives. The Liberal lead with men in Ontario is much tighter, 39.8 to 37.6.

Nationally, the older you are the more likely you are to vote Conservative. The Liberals hold the advantage up to age 44, then the Conservatives the rest of the way. Interestingly, while the Conservatives still lead, as people leave middle-age for their golden years, there’s an uptick in Liberal support, which indicates targeting seniors would be advisable.

On the education front, don’t think I’m making any judgement here but the more education you have, according to Ekos you’re more likely to vote Liberal. Those with university degrees are LPC 38.9 to CPC 30.2, but those with high school or college favour the CPC by five to six points.

Beyond the surface

Some of the additional questions that Angus Reid asked help to illustrate a point I like to often make, and that’s that numbers, on the surface, can be deceiving.

For example, Angus Reid reports that 41% oppose toppling the government and 16% moderately oppose it, with just 14% strongly supporting it and 18% moderately supporting it. Regionally, opposition is strongest in Manitoba and Saskatchewan at 61% strongly opposed, and support strongest in Atlantic Canada at 25% strongly support.

So, we can take from those numbers that an election is a bad idea and the perceived culprit will be smacked for it, right?

Well, maybe not.

When asked the change question (it’s time for a new party in government, or it’s not change the CPC should stay), 42% said it was time for change while 34% wanted the CPC to stay. Of course, those 42% will disagree on what the change should be, naturally.

Still, the contradiction between the opposition to election and desire for change is there. What does it mean? I’d argue it means that, while Canadians aren’t keen on the idea of election, that’s not surprising because they never are. But they’ll still go and vote, they’re looking for change, and there is unlikely to be any kind of election backlash, no matter how hard the spinmiesters try.

Anyway, while interesting when a new dynamic unfolding now these polls mean nothing. They’re merely the starting-point, not the destination.

Recommend this Post on Progressive Bloggers


Robert said...

You do know the NDP has 2 seats in the GTA right?

Jeff said...

I think I heard something about that Robert, yes. And at 11.8 per cent, they seem unlikely to do better. Indeed, they'd better hope all those people like in Jack and Olivia's ridings.

Robert said...

They only got around 12-14% in the GTA last time. Too bad Bob Rea can only win Toronto centre once eh? These numbers aren't good news but they aren't a move away from the status quo either.